A critical review of mid- to late-Holocene sea-level ... · A critical review of mid- to late-Holocene sea-level ﬂuctuations on the eastern Brazilian coastline Rodolfo J. Anguloa,,

ARTICLE IN PRESS

0277-3791/$ - se

doi:10.1016/j.qu

�CorrespondE-mail addr

[email protected]

Quaternary Science Reviews 25 (2006) 486–506

A critical review of mid- to late-Holocene sea-level fluctuationson the eastern Brazilian coastline

Rodolfo J. Anguloa,�, Guilherme C. Lessab, Maria Cristina de Souzaa

aDepartamento de Geologia, Universidade Federal do Parana, Cx. Postal 19001, Curitiba (PR) 81531-970, BrazilbInstituto de Geociencias, Universidade Federal da Bahia, Campus Ondina, Salvador (BA) 40170-280, Brazil

Received 13 June 2004; accepted 3 March 2005

Abstract

The eastern coast of Brazil is 7000 km long and has been the subject of systematic paleo-sea-level investigations for more than 35

years. More than 1000 samples have been radiocarbon dated, and paleo-sea-level trends have been determined for 14 coastal sectors.

These trends have in common a mid-Holocene sea-level maximum (PMT) above present sea-level and a subsequent fall to the

present time. The time and elevation of the PMT, the time when relative sea-level rose above present mean sea-level, and the nature

of late Holocene sea-level fall can, however, differ significantly. Discrepancies are observed not only between neighboring coastal

sectors, but also between different studies in a same coastal sector. This paper discusses all the key radiocarbon-dated samples used

to establish regional trends in relative sea-level on the eastern coast of Brazil. It is concluded that although many of the key sea-level

indicators are imprecise (in time and space), there is widespread evidence (30% of the data set) for a progressive decline, possibly

with uneven rates, of relative sea-level since the end of the PMT.

r 2005 Elsevier Ltd. All rights reserved.

1. Introduction

The eastern Brazilian coastline, between the parallels51S (State of Rio Grande do Norte) and 341S (State ofRio Grande do Sul), is about 7000 km long andencompass about 60% of the Atlantic South Americancoast (from Argentina to Trinidad and Tobago)(Fig. 1a). Systematic studies of the paleo-sea-level inthis region started in the mid-1960s, with more than 100publications addressing the Holocene sea-level historypublished since then. These studies show that a mid-Holocene sea-level highstand occurred along the wholecoast, followed by a drop to present time. However,several aspects of the relative sea-level history areproblematic and debatable, given the poor qualityof many paleo-sea-level indicators combined with

e front matter r 2005 Elsevier Ltd. All rights reserved.

ascirev.2005.03.008

ing author

esses: [email protected] (R.J. Angulo),

ba.br (G.C. Lessa).

divergent approaches to their interpretations. Criticalareas of debate during the last 7000 years include thedifferent elevation of the highstand and the presence orabsence of high-frequency sea-level oscillations. Resol-ving these issues is important since a well-tuned paleo-sea-level curve, or envelope, is necessary for the correctunderstanding and assessment of coastal and oceanprocesses, such as rates of sedimentation, net sediment-transport volumes and coastal erosion or progradation.It is also essential for the correct development ofevolutionary models of coastal landscapes (barriers,estuaries and coral reefs), as well as the calibration ofisostatic models.

This paper presents a comprehensive review of thepaleo-sea-level indicators used to date to develop a newmodel of the Holocene sea-level trend along the easternBrazilian coast. To achieve this we will review: (i) thetiming when sea level first reached present mean sea-level (MSL) during the postglacial sea-level rise; (ii) themaximum elevation of sea level in the mid-Holocene

ARTICLE IN PRESSR.J. Angulo et al. / Quaternary Science Reviews 25 (2006) 486–506 487

and when it occurred; (iii) the nature of sea-level fall inthe late Holocene and; (iv) the evidence for significantregional differences in the Holocene sea-level record.The work synthesizes a large amount of informationon paleo-sea-level indicators previously published inBrazilian journals and conference proceedings ofrestricted circulation.

(a)

Fig. 1. (a) Location of the sites from where dated paleo-sea-level indicators h

in the coastal sectors of (b) Salvador, (c) Cananeia-Iguape, (d) State of Para

2. Chronology of relevant paleo-sea-level studies in Brazil

To help the understanding of the present debate onthe Holocene Brazilian sea-level history, it is importantto provide an insight on the succession of articles thatled to the widely held Brazilian sea-level curve that isnow the subject of this review.

ave been obtained along the eastern Brazilian coast. Insets show details

na and Itapoa, (e) Ilha de Santa Catarina and (f) Laguna.

ARTICLE IN PRESS

(f)(d)

(b) (c)

(e) (f)

Fig. 1. (Continued)

R.J. Angulo et al. / Quaternary Science Reviews 25 (2006) 486–506488

The first references to Holocene paleo-sea-levels inBrazil date back more than 100 years, with the works ofHartt (1870) and Branner (1902). In the 1940s, moredetailed studies of the Brazilian sea-level oscillationswere published by Lamego (1940), Bigarella (1946),Maack (1947, 1949) and Almeida (1955). In the 1960s,the first radiocarbon dates of biogenic limestonecomposed mainly by vermetids (Van Andel and Laborel,1964), allowed for the earliest reliable temporal andspatial interpretations of paleo-sea-levels. Other pioneerradiocarbon dates for paleo-sea-level studies were madeby Bigarella (1965, 1971), Bigarella and Sanches (1966),Delibrias and Laborel (1969) and Laborel (1969).

Bigarella (1965, 1971) compared the paleo-sea-level datafrom Brazil with Fairbridge’s (1961) curve, which wasunderstood at that time as a representation of theeustatic sea-level behavior. Delibrias and Laborel (1969)were the first to anticipate regional differences in thesea-level history, showing that, in contrast to sites in theNorthern Hemisphere, a clear highstand at least 1mhigher than present existed in Brazil.

Pioneer outlines of the sea-level curves (from Cana-neia coastal plain and Baıa de Paranagua—Figs. 1c, d, 2and 3), heavily laden with archeological information,were presented in the beginning of the 1970s by Martinand Suguio (1975) and Suguio et al. (1976). It is already

ARTICLE IN PRESS

(a)

(b)

Fig. 2. (a) Sea-level changing scheme for the last 6000 years in Cananeia-Iguape region and (b) sea-level changing indicated by the shell-midden from

Ilha das Rosas in Baıa de Paranagua (after Martin and Suguio, 1975). For location see Figs. 1c and d and details on the identified sample see Table 1.

(a)

(b)

Fig. 3. Relative sea-level changes in the Cananeia-Iguape region: (a) information from geomorphologic and geological evidence and (b) additional

information supplied by samples from shell-middens (after Suguio et al., 1976). For location see Fig. 1c.

Fig. 4. Relative sea-level changes in the last 7000 years in the region around Salvador (after Martin et al., 1979/1980). For location see Fig. 1a and for

details on the identified samples see Table 1.

R.J. Angulo et al. / Quaternary Science Reviews 25 (2006) 486–506 489

observed in these curves the general shape of thecommonly accepted Brazilian relative sea-level curve,with shell-midden indicating the periods of lower sea-level and the elevation of marine terraces indicatingperiods of higher sea-levels.

In the late 1970s, Martin et al. (1979/1980) publisheda curve that became the reference of the BrazilianHolocene sea-level behavior, with the two high-fre-quency oscillations and the highstand at 4.8m abovepresent MSL (Figs. 1a and 4). It is important to note,

ARTICLE IN PRESSR.J. Angulo et al. / Quaternary Science Reviews 25 (2006) 486–506490

however, that only six paleo-sea-level indicators, out ofthe total 52 used, have a clearly defined age and heightrelationship to a former sea level (see below).

In the mid-1980s, Suguio et al. (1985) compiled allpublished curves between the states of Alagoas andSanta Catarina (Fig. 1a), that in accordance with theauthors, were the result of more than 700 14C dates. Thepapers by Martin et al. (1979/1980) and Suguio et al.(1985) were complementary, and grounded the conceptof the high-frequency sea-level oscillations that hasbiased the interpretation of the coastal geomorphologyin the country (see discussion below).

Other relevant studies discussing paleo-sea-level dataand trends (Villwock et al., 1986; Dominguez et al.,1990; Angulo and Suguio, 1995; Angulo and Lessa,1997; Lessa and Angulo, 1998; Tomazelli et al., 1998;Angulo et al., 1999) were published in the late 1980s and1990s. With the exception of the work by Dominguezet al. (1990), all the other publications proposedalternative interpretations for the sea-level data andtrends. For instance, Tomazelli et al. (1998) suggestedthat sea level has been rising in the last 2000 years,Angulo and Suguio (1995) questioned whether the geoidcould cause the difference in the elevation of thepostglacial sea-level maximum (PSLM) along the coast(as proposed by Martin et al., 1985), and Angulo andLessa (1997) and Lessa and Angulo (1998) contested theexistence of the high-frequency sea-level oscillations andthe elevation of the Holocene sea-level maximum.

More recently Bezerra et al. (2003), based on 48radiocarbon dates, showed that regional tectonics islargely controlling the relative sea-level variations in theState of Rio Grande do Norte (Fig. 1a). Finally, Martinet al. (2003) summarized 25 years of studies sustainingthe sea-level behavior indicated by the base curve. Inaddition, corrections to account for the fluctuation ofatmospheric CO2 through time was performed, and thecurves were then plotted in calibrated years.

3. The data set

A considerable amount of contradictory informationexists regarding the samples used for sea-level studies inBrazil, including their location, nature, ages, laboratoryreference number and inferred paleo-sea-levels. Thisrequired an extensive search in more than 100 articlesand technical reports where radiocarbon dates wereoriginally published and first described. Wheneverpossible, contact was made with the authors or withthe laboratory responsible for the radiocarbon datings.Some of the original data have, therefore, beencorrected, and differ from those published elsewhere.Table 1 lists the paleo-sea-level data set, excludingsamples older than 28,000 yr BP as well as samplesobtained from archaeological debris (shell-middens). All

dates are calibrated after Stuiver and Reimer (1993),with calibrated dates using a two sigma age range. Aregional reservoir effect of 8717 years, determined byAngulo et al. (2005) for the south and southeasternBrazil, was attributed to all marine samples.

4. Discussion

Different paleo-sea-levels can be generated by (i)misinterpretations of the elevation of the paleo-sea-level,(ii) misinterpretations of the age of the paleo-level, (iii)leveling errors, (iv) dating errors caused by contamina-tion, and (v) inclusion, or not, of time correction factors,such as carbon isotopic fractionation, variation ofatmospheric 14C content through time and reservoireffects. Misinterpretations of the paleo-environment ofdeposition, misunderstanding of the vertical zonationof marine species, and the use of inappropriate orinadequate evolutionary models of coastal landscapescan also result in an incorrect reconstruction of paleo-sea-level. Age misinterpretations commonly occur wheninformation on the paleo-sea-level and its time comefrom different sources (Angulo et al., 2002). An exampleis a beach-face sediment deposit with embedded shells.Dating of the shells provides a time control and theenvironment of deposition provides an elevation con-trol. However, it often occurs that the age of reworkedsamples is taken as the age the sedimentary deposit fromwhich it is derived.

The following sections review the main areas ofdebate in Brazilian sea-level research that, as willbecome apparent, revolve to a large degree around theidentification and treatment of these various age andaltitude issues.

4.1. When did relative sea-level reach present mean sea-

level in the postglacial transgression?

The timing of when relative sea-level rose abovepresent MSL provides a useful datum to assess whetherlocal or regional differences in the relative sea-levelhistory exist in Brazil. Suguio et al. (1985), for example,noted that there appeared to be a northward delay of thetime when this crossover occurred. This delay was alsonoted by Martin et al. (1985), who ascribed it tovariations in the configuration of the geoid over Brazilduring the late Holocene. Between the states of Rio deJaneiro and Parana, where the coast is oblique to thegeoid contour lines, a slight east–west geoidal shiftwould lower the elevation of the PSLM and diminish theamplitude of the subsequent sea-level fall (Fig. 5). Alsoas a consequence, the crossover time between Rio deJaneiro and Parana would occur later than in the otherscoastal sectors (Figs. 4 and 6).


Although the age estimates provided for the relativemean sea crossing times appear precise, there are majorage and altitude uncertainties associated with theseestimates. These arise from the nature of the materialdated and also from the interpretation of the paleo-environment from which these critical indicators are

(a)

(b)

Fig. 5. (a) Geoidal chart of Brazil and (b) a cross section of the geoid

today and a suggested one for the mid-Holocene with a lateral offset.

Regional differences in the elevation of the postglacial sea-level

maximum were ascribed to subsidence of the geoidal relief east of

Cananeia, which was due to a slight displacement of the central

depression (after Martin et al., 1985). The location of the profile in (b)

is showing in (a).

Fig. 6. Paleo-sea-level curves for three different sectors of the Brazilian coast

location see Fig. 1c.

derived. This is now illustrated through detailedconsideration of the relevant data from two sites.

The MSL-crossing times have been established byMartin et al. (1985) and Suguio et al. (1985) at 6600 yrBP in Cananeia-Iguape (251S), 6800 yr BP in Santos(241S) and 7100 yr BP in Salvador (131S) (Fig. 1a). Inearlier papers, the MSL-crossing time in Cananeia wasgiven as 6200 yr BP (Suguio et al., 1976) and 6900 yr BP(Martin et al., 1979b), whereas in Santos an age of6400 yr BP (Martin et al., 1979b) was given. Two othersamples from a costal plain near Santos (Praia Grande),although indicating paleo-sea-levels higher than present,can also be used to infer an approximate time for theearliest MSL-crossing time, and will be discussed below.

In Cananeia-Iguape and Santos, the MSL-crossingtime was determined by interpolating between the agesof different samples. The samples consist of transportedwood fragments within muddy sediments (Suguio et al.,1976; Martin et al., 1979b). One was dated at65007170 yr BP (Bah-230, 7659–6949 cal yr BP), andwould indicate a paleo-sea-level of �0.870.3m (Suguioet al., 1976) (Fig. 7a). The other sample provided an ageof 61007130 yr BP (Bah-226, 7248–6486 cal yr BP), andwould indicate a paleo-sea-level of 0.870.3m (Fig. 7b).Other similar indicators found in the states of Sao Paulo

between Angra dos Reis and Cananeia (after Martin et al., 1979b). For

(a)

(b)

Fig. 7. Schematic cross sections of the sites where radiocarbon-dated

samples were collected: (a) landward margin of Ilha Comprida and (b)

Canal de Ararapira (after Suguio et al., 1976). For location see Fig. 1c.


and Rio de Janeiro support this MSL-crossing time(Martin et al., 1979b).

The problem with the determination of the MSL-crossing times above resides on the interpretation of thedepositional environment. This interpretation was basedon a flawed conceptual model of the evolution of thecoastal plain (see Angulo and Lessa, 1997). Forexample, it is possible that the muddy sediments fromwhere the samples were collected were indeed a paleo-estuarine deposit, but of sub-tidal elevation. This wouldindicate the lowest possible sea level at the time. It isalso possible that the age of the deposit is much youngerthan the (transported) sample. In this case, bothelevation and age controls are inadequate to indicatethe MSL-crossing time.

The MSL-crossing time in Salvador was based on asingle sample: a reworked shell embedded in a beach-rock indicating an elevation of 070.5m at 70957125 yrBP (Bah-571, 7792–7334 cal yr BP) (Bittencourt et al.,1978; Martin et al., 1979a). This altitude interpretationwas based on primary sedimentary structures observedin the beach-rock. The authors inferred the age of thedeposit by the age of transported samples, which can bemuch older than the deposit itself. In addition, Lessaand Angulo (1998) suggested that the depositionalenvironment should be at least 1.5m higher (see section4.3.2). Therefore, problems associated with time andelevation control of this sample makes it also unsuitablefor the determination of the crossover time in the regionof Salvador.

The only in situ paleo-sea-level indicator in the wholedata bank is the mangrove trunk found in living positionin a trench dug on the coastal plain close to Santos(Fig. 8) (Martin and Suguio, 1978; Martin et al., 2003).The trunk was rooted into muddy sediments and buried

Fig. 8. Schematic profile of the Holocene terrace in Praia Grande, State of Sa

with ages older than 6800 cal yr BP were collected. The profile also shows fou

For location see Fig. 1c.

by marine sands, suggesting either a transgressive eventor lateral barrier progradation. It rested 2.070.5mabove the level where modern mangrove trees grow andwas dated to 6480775 yr BP (Bah-327, 7557–7264 cal yrBP) and 62507130 yr BP (Gif-3845, 7423–6803 cal yrBP) (Fig. 8). By incorporating these two paleo-sea-levelreconstructions in the sea-level base curve of Salvador,Martin et al. (2003) suggested an MSL-crossing time ataround 7550 cal yr BP.

It can be concluded that complex indicators indicatethat sea level had already overtaken present MSL by6600 yr BP (6900–7700 cal yr BP). However, the majorityof these indicators are not sufficiently precise toestablish regional differences in the MSL-crossing time.Consequently, there is also no indication that changes ofthe geoidal surface have occurred, as proposed byAngulo and Suguio (1995).

4.2. The mid-Holocene highstand

References to the elevation of the mid-Holocenehighstand in the eastern coast of Brazil can be foundin several publications, including Suguio et al. (1985),Dominguez et al. (1990), Angulo and Suguio (1995),Angulo and Lessa (1997), Angulo et al. (1999), Souzaet al. (2001) and Bezerra et al. (2003). The suggestedelevations vary from a minimum of 2.171.0m inLaguna (Angulo et al., 1999) to 5m in Pernambuco(Dominguez et al., 1990) (Fig. 9). Suguio et al. (1985)suggested regional differences of 2.5m amongst sevensites between Salvador in the north and Laguna in thesouth, and ascribed this difference, as stated in theprevious section, to a shift of the geoidal relief (Martinet al., 1985; Suguio et al., 1985) (Fig. 5).

o Paulo showing the location of the tree trunk from where two samples

r other dated samples listed in Table 1 (after Martin and Suguio, 1978).

ARTICLE IN PRESS

Fig. 9. Elevation of the Holocene highstand along the eastern coast of Brazil in accordance with different authors.


The time of the highstand varies between 51957110 yr BP (Bah-567, 5783–5291 cal yr BP) (Martin et al.,1979a; Martin et al., 2003) and 5410780 yr BP (CENA-179, 5916–5597 cal yr BP) (Angulo et al., 1999). All ofthe above ages and elevations were obtained with theelevation of coastal barriers, dating of oyster samples,and shell-rich sediment layers on beaches, that will benow discussed.

4.2.1. Elevation of the mid-Holocene highstand inferred

through the elevation of coastal barriers

Martin and Suguio (1978), Suguio and Martin (1978),Martin et al. (1979/1980, 1988a), Suguio et al. (1985)and Martin et al. (2003) suggest that the elevation ofwave-built terraces (beach ridges) represent the MSL atthe time of their deposition. They use this assumption tohypothesize the Holocene sea-level maximum in thecoastal sectors between Itacare and Laguna (Fig. 1a).Dominguez et al. (1990) also infer the elevation of theHolocene sea-level maximum on the basis of theelevation of a wave-built terrace.

As Angulo and Lessa (1997) pointed out, theelevation of the terraces are related to the elevations ofthe swash limit during storms, which in its turn dependson the wave spectrum, beach grain size and slope, thebathymetry of the nearshore and the tidal range. Roep(1986) showed that paleo-sea-levels in the Netherlands’scoast should be 2–3m below the surface of the wave-built terraces. Bigarella et al. (1961) showed that wave-built terraces with different elevations are associatedwith the present MSL along the coast of the State ofParana, and indicated that the PSLM may have been1–6m below the average elevation of the inland-mostpart of the strandplain (Angulo, 1994). Therefore, we

consider that all the above paleo-sea-level interpreta-tions based on the elevation of wave-built terraces areoverestimated, whereas their margin of error is sig-nificantly underestimated.

In the Paranagua region (Fig. 1a), Suguio et al. (1985)and Martin et al. (1988b) inferred the elevation of theHolocene highstand on the basis of the altitude ofinnermost part of the strandplain without indurated,organic-rich, dark-brown sands. It was then assumedthat epigenetically enriched sediments were character-istic of Pleistocene deposits only. However, Angulo andSuguio (1995) and Angulo et al. (2002) showed that suchindurated sands can also be Holocene. Angulo andSuguio (1995) re-evaluated the morphological boundarybetween the Pleistocene and Holocene strandplains, andthe elevation of the highstand was then re-assessed atabout 4m (Fig. 10).


with shell-rich sediment layers

Inferences of the paleo-sea-level made by Martin andSuguio (1976, 1978, 1989), Suguio and Martin (1978)and Martin et al. (1979b, c, 2003) based on shell-richsediment layers are questionable due to the inaccurateinterpretation of the depositional environment.

For instance, close to Santos (Morro da Enseada,Fig. 1c), in a protected re-entrance of the coastline, theelevation of the transgression maximum has beeninferred from shell-rich layers embedded in beach-rocks.Two samples obtained from the shelly layers were datedat 54707100 yr BP (Bah-609, 6070–5600 cal yr BP)and 3475770 yr BP (Bah-355, 3512–3162 cal yr BP)(Fig. 11). Martin and Suguio (1976) and Martin et al.(1979c) suggested that the shell-rich layers are analogous

ARTICLE IN PRESS

(a)

(b)

Fig. 11. (a) Schematic profile of the Morro da Enseada outcrop in de

Bertioga coastal plain with the location of the samples and (b) the

position of the samples in relation to the Santos paleo-sea-level (after

Martin et al., 1979c). For location see Fig. 1c and for details on the

identified samples see Table 1.

Fig. 10. Heights of the maximum relative sea-levels along the State of Parana coastline during the last 6000 yr BP. The original curve for Parangua

from Suguio et al. (1985) was elevated based on re-interpretation paleo-sea-level and the age of the coastal plain (after Angulo and Suguio, 1995). For

location see Fig. 1a.


to similar deposits found in the lower half of the beachprofile, between MSL and low-tide level. The elevationdifference between the shell layers on the present beachand on the outcrop suggests a sea-level maximumbetween 4.070.5 and 4.670.4m (Martin and Suguio,1976, 1978; Suguio and Martin, 1978; Martin et al.,1979c, 2003). Because the age 3475770 yr BP was tooyoung to be associated to the PSLM, and the d13C wasrather negative (�7.25%), Martin et al. (1979c) con-sidered that the sample was contaminated (Fig. 11).

Martin et al. (1979c, 2003) interpreted the sedimen-tary sequence in the outcrop as transgressive in nature,suggesting that the top of the terrace was an environ-ment analogous to the lower part of the present-daybeach. The lower beach would have been elevated by thevertical accretion of a low-tide terrace. This is a ratheruncommon coastline adaptation to a rising sea level,with no landward displacement of the shoreline indeed.If shoreline migration was impaired by the presence of acoastal relief with higher topographical gradients, the

morphological result would more likely be a drownedsandy body (Roy et al., 1995). Although we have notvisited the area, we consider that the shell layer inquestion more probably represents an elevated part apocket beach, and not an environment analogous to thelower beach face. Any inference of the paleo-sea-levelusing this evidence is therefore highly questionable.


from oyster samples

Because oyster can live in a wide range of elevation,they are of limited use as a paleo-sea-level indicator.Oyster samples were used to determine the elevation ofthe PSLM close to Angra dos Reis-Parati, Rio deJaneiro (Fig. 1a). Two in situ samples dated at52007200 yr BP (LJ-1364, 5982–5019 cal yr BP) and48007200 yr BP (LJ-970, 5558–4558 cal yr BP) (Martinand Suguio, 1978) indicate an elevation of 4.8m. Thissample was collected and dated by Danciger and Curray(pers. comm. in Delibrias and Laborel, 1969), whomeasured the samples’ elevation as 4.8m above MSL.As Laborel (1979, 1986) pointed out, the verticalzonation of a rocky cliff does not correspond to thehydrographic tide levels. Delibrias and Laborel (1969)observed ‘‘living Ostrea in wave beaten crevices, 3mabove mean above mean level’’ at Cabo SantoAgostinho (Fig. 1a), and concluded that oysters areparticularly poor sea-level indicators and that, as aresult, it is possible that the level given by Danciger andCurray is an overestimate. Therefore, the elevation ofthe PSLM around Angra dos Reis could have beenlower than 4.8m.


from vermetid samples

The most reliable inferences on the elevation of theHolocene highstand (Fig. 9) are given by the vermetids.In the region around Salvador, vermetid samples werecollected at the entrance of Baıa de Todos os Santos(Porto da Barra headland, 131S, Fig. 1b), providing anage of 51957110 yr BP (Bah-567, 5783–5291 cal yr BP)


and an inferred paleo-sea-level of 4.770.5m (Martinet al., 1979a, 2003) or 4.570.5m (Martin et al., 1979d)or 4.870.5m (Suguio et al., 1985) (Table 1). Martin etal. (1979d) mentioned that the assessed paleo-sea-levelcould be slightly overestimated, because the site wasexposed to the wave action. Nevertheless, the 4.770.5melevation is maintained by Martin et al. (2003) becausealthough it ‘‘could at first be a source of controver-syyinformation from other sectors of the Braziliancoast corroborates the timing and/or altitude of themaximum of the Holocene Transgression in theSalvador curve’’.

It is worth pointing that at least 142 paleo-sea-levelassessments have been published on the basis ofvermetid (138 samples) or algae with vermetids (foursamples) samples, amongst which only two, besides thesample from Salvador, indicated paleo-sea-levels higherthan 4m. A sample obtained in Gaibu provided an ageof 38707170 yr BP (Bah-1221, 4277–3397 cal yr BP) andindicated a paleo-sea-level of 4.8370.5m (Dominguezet al., 1990). The other, collected at Cabo de SantoAgostinho was dated at 20107160 yr BP (Bah-1236,1931–1246 cal yr BP) and indicated a paleo-sea-level of4.3670.5m (Dominguez et al., 1990). We consider thatthese samples are exceptions, possibly due to the highwave action at the site and, as will be shown below, areoutliers within the data set.

Another four vermetid samples have been used toestablish the maximum PSML in the states of Paranaand Santa Catarina. In Morro de Caioba (Fig. 1d),Angulo et al. (2002) collected two samples at anelevation 3.671.0 and 3.571.0m above the level ofoccurrence of the polichaete Phragmatopoma lapidosa,whose upper living zone corresponds to the lower livinglimit of the vermetid Petaloconchus (Macrophragma)varians (vermetids are no longer found in southernBrazil—Laborel, 1977; Angulo et al., 1999). The samplesprovide an age of 4750770 yr BP (CENA-140,5222–4826 cal yr BP) and 5300770 yr BP (CENA-141,5845–5513 cal yr BP), in agreement with the elevationsuggested by Angulo and Lessa (1997) based on theelevation of the outermost part of the Pleistocenestrandplain. In the State of Santa Catarina, two othervermetid samples were obtained by Angulo et al. (1999)from rocky headlands between Cabo de Santa Martaand Ponta de Itapiruba (Fig. 1f), at elevations of2.171.0 and 1.9571.0m. The dating of these samplesprovided ages of 5410780 yr BP (CENA-179,5916–5597 cal yr BP) and 4600770 yr BP (CENA-192,4971–4577 cal yr BP).

All the evidence presented by the indicators abovepoint out that the PSLM took place between 4900 yr BP(5100 cal yr BP) and 5400 yr BP (5700 cal yr BP), andreached an elevation between 3 and 4m from the state ofPernambuco to the state of Parana, and 2m in thesouthern State of Santa Catarina.

4.3. Late Holocene high-frequency oscillations of sea

level

The existence of the high-frequency sea-level oscilla-tions in Brazil after the PSLM has long been debated(Angulo and Lessa, 1997; Martin et al., 1998; Lessa andAngulo, 1998; Martin et al., 2003). It is hypothesizedthat these fluctuations had amplitudes of 3–4m, andoccurred between 4100–3800 and 3000–2700 yr BP(4200–3700 and 2700–2100 cal yr BP) (Suguio et al.,1985; Martin et al., 2003).

Evidence for the regressive events were presented inthe first sea-level curve for the Cananeia-Iguape regionpublished by Martin and Suguio (1975) (Fig. 4a). In thispaper, two lines of evidence point to lower sea-levelsbetween 4100 and 3800 yr BP (4200–3700 cal yr BP): theshell-midden at Canal do Ararapira (Fig. 2a) and asample of plant detritus obtained from an outcrop closeto Barra de Icapara (Fig. 2a). Another shell-midden,located in Ilha das Rosas (Baıa de Paranagua), providesevidence of a second oscillation (Fig. 2b).

Several indicators were subsequently presented tosupport the existence of these regressive events, and canbe grouped into five classes: (a) shell-middens (from theregions of Salvador, Santos, Cananeia-Iguape and Baıade Guaratuba); (b) primary sedimentary structures inbeach-rocks (from Salvador region); (c) detritus ofbeach vegetation (from Cananeia-Iguape region); (d)swamp deposits (from Rio do Fogo) and (e) truncationof beach ridges within the strandplains at Caravelas andat the mouth of the rivers Doce, Jequitinhonha andParaiba do Sul. Amongst these, only the shellsembedded in beach-rocks provide good-quality spatialand temporal control on former sea-level.

4.3.1. Evidence of lower sea-levels derived from shell-

middens

Shell-middens are the most numerous and perhaps thesecond most important evidence, following beach-rocks,of the short-term regressive episodes, and have beenused extensively in the work of Martin and Suguio(1976, 1989, 1992), Martin et al. (1979a, b, 1986a, b,1987, 1988a, b, 1996, 1997, 1998, 2003) and Suguio et al.(1980, 1985, 1986, 1992). These middens, especially insouth Brazil, can be as long as 300m, as wide as 60mand taller than 30m (Martin et al., 1986a, b; Gaspar,2000). The potential of shell-middens for sea-levelstudies was anticipated by Laming-Emperaire (1968),and thoroughly discussed by Martin et al. (1986a, b).Laming-Emperaire (1968) mentioned the existence ofseveral shell-middens that extend below presentsea-level, and noted specifically the shell-middens ofIlha das Rosas (in Baıa de Paranagua, Fig. 1d) andMaratua (close to Santos, Fig. 1c). Martin and Suguio(1975) and Martin et al. (1979a, 1986a, b, 2003), besidesthese middens, refer to others found at Canal de

ARTICLE IN PRESS

(a)

(b)

Fig. 12. Detailed scheme of internal structures of the shell-midden of

(a) Ponta das Almas and (b) Carnic-a I (after Hurt, 1974). For location

see Figs. 1e and f, respectively.


Ararapira (in the Cananeia-Iguape region, Fig. 1c), atPedra Oca (in the margins of Baıa de Todos os Santos,Fig. 1b) and at Rio Boguac-u (inside Baıa de Guaratuba,Fig. 1d), as evidence of paleo-sea-levels at or below thepresent level.

The two assumptions made by Laming-Emperaire(1968), Martin and Suguio (1976) and Martin et al.(1986a, b, 1987, 2003) in the use the middens as paleo-sea-level indicators are that they are a remnant, or wastedumps, of campsites, and as such, their base (i.e., itsinitial stages) would have to be originally above thespring high-tide level at the time of construction. Anguloand Lessa (1997) questioned the validity of the historicsea-level data obtained from shell-middens, given thevarious possibilities that could have led an unknownculture to initiate a mound underwater. Recent arche-ological studies (Gaspar, 1998, 2000; Fish et al., 2000;Nadal de Masi, 2001; Gaspar et al., 2002) indicate thatshell-middens were structures with multiple functions,with a complex and hierarchical spatial distribution thatresulted from a socially planned effort. Amongst others,one of the reasons for erecting a midden would be thecreation of an important topographical reference in thelandscape. Gaspar (1998) reported that some middenswere used as living sites, others were dumpsites and stillothers were burying grounds. Hence, although initiallysound, the idea that shell-middens are the daily liferemains of pre-historic people accumulated on dryground must be re-evaluated.

Furthermore, the nature of the midden’s substrateand the midden’s complex internal structure bring inadditional complexity to the assessment of the originalsubstrate elevation and paleo-sea-levels. Detailed strati-graphic studies of shell-middens reveal a highly complexinternal structure, making it difficult to assess where theearliest layers are. The shell-middens of Ponta dasAlmas and Carnic-a I, studied by Hurt (1974) in thecoast of Santa Catarina (Fig. 12), are good examples ofsuch complexity. A detailed investigation of theirinternal structure reveal several discontinuous timelines, suggesting the merging of different smallermounds that were initiated over different substrates(Fig. 12). Also, as the mound grew, slumping wasprobably a natural process. This indicates that, even ifthe original premises taken by Laming-Emperaire(1968), Martin and Suguio (1976) and Martin et al.(1986a, b, 1987, 2003) were correct, investigation andsampling of numerous detailed sections in the samemidden are needed to identify its base. We have foundno indication that any further evaluation of the shell-midden structures has been performed to assess theirsignificance as paleo-sea-level indicators.

The d13C of the mollusk shell samples obtained fromshell-middens have also been used as evidence of lateHolocene sea-level oscillations. Because d13C values arehigher in marine than in continental waters, the value

recorded in mollusk shells is an indication of the degreeof marine (continental) influence in the paleo-environ-ment (Martin et al., 1986a, b, 2003). Decreasing(increasing) value of d13C has then been interpreted asa consequence of a fall (rise) of sea level, and a higher(smaller) degree of continental influence upon theestuary. Apart from possible relative sea-level changes,the degree of marine influence inside an estuary is alsoregulated by natural changes in the estuarine morphol-ogy during the sediment-infilling process (that canchange the tidal range and excursion), and by naturalfluctuations of the freshwater inflow due to climaticoscillations. Regarding the latter, Genz et al. (2003) andPekarova et al. (2003) have shown that low-frequency(�30 years) hydrological cycles are associated withchanges in the mean annual river discharge of up to 40%(as it is the case for rivers in the State of Bahia—Genzet al., 2003). These cycles can certainly impact theestuarine ecology, as indicated by Alber (2002). Theshells collected by the primitive people belonged mostlyto the species Anomalocardia brasiliana, Ostrea, Cras-

sostrea and Modiolus (Bigarella, 1950/51a, b; Emperaireand Laming, 1956; Laming-Emperaire, 1968; Hurt,1974; Gaspar, 1998, 2000). The life-time of A. brasiliana

varies between 1 and 3 years (Boehs, 2000) and that ofOstrea and Crassostrea is around 20 years (Wakamatsu,1973). This means that the life cycle of an estuarinemollusk may record short-lived alternation betweendrier and wetter periods and that the simple assumptionthat d13C values can be used to reconstruct changes inrelative sea-level is open to doubt.

ARTICLE IN PRESS

(m)

(m)

MSL

4

3

2

1

0

0 20 40 60 80 100 120 140

Fig. 14. Cross section of the beach-face south of Arembepe and beach-

rock exposures with their primary sedimentary structures identified.

The lower beach-rock has trough cross-bed structures, whereas the

upper beach-rock, with tabular beds of conglomerates interfingering

with plane-parallel laminations of sand (after Lessa and Angulo, 1998).

For location see Fig. 1b.


4.3.2. Evidence of lower sea-level derived from primary

sediment structures in beach-rocks

Six samples (Bah-589, Bah-497, Gif-2150, Bah-617,Bah-525 and Bah-496—Table 1) of bivalve shellsobtained from beach-rocks in the northern coast ofBahia also were used as evidence for the secondaryoscillations of the paleo-sea-level (Martin et al. (1979/1980). Radiocarbon dating of reworked shells provideda time control, whereas primary sedimentary structureswithin the beach-rocks defined the environment ofdeposition, with their elevation relative to a presentanalog defining the paleo-sea-level (Martin et al., 1979a;Flexor and Martin, 1979). These six samples constitutedthe only evidence, with both time and space controls, oflower sea-levels in the proposed intervals for thesecondary oscillations. Martin et al. (2003, Fig. 7 page107) included another three, non-referenced, samples inthe 4100–3800 yr BP time interval, besides removingone sample from the time interval between 2900 and2700 yr BP.

Based on the textural characteristic of the deposit(a conglomerate, as will be shown below), Flexor andMartin (1979) and Martin et al. (1979a) suggested thatdeposition occurred in the lower part of the higherbeach (Fig. 13). The authors mention that the beach-rock is located approximately at the level of the changein slope on the modern beach (beach step), thereforeindicating a similar environment of deposition, or apaleo-sea-level close to the present one. Flexor andMartin’s (1979) interpretation of the environment ofdeposition of the dated shells was based on amorphosedimentary model that sub-divided the beachin four zones, accounting for variations in sedimenttexture and sedimentary structures (Fig. 13). This modelshows the same lateral variation of the beach structuresas the Reineck and Singh’s (1973) model, althoughchanges in the sedimentary structures occur at differentelevations along the profile. For example, the transitionbetween plane-parallel and cross-stratified laminationsoccurs above the low water level in the model of Flexorand Martin (1979), whereas it is below low water level inthe model of Reineck and Singh (1973). Cross-stratifiedlaminations are rather difficult to form on the beach

Fig. 13. Schematic beach profile and variations in the sedimentary st

face, since supercritical flows of the wave swash tend todestroy any cross-stratified laminations produced by ashort lived (half tidal cycle), weaker longitudinal flow(see, e.g., Reineck and Singh, 1973; Short, 1984).

In the field site, we have observed two beach-rockswith different matrices on the beach face (Fig. 14). Thelower beach-rock is observed from about MLWS toMHWS, and is composed of fine and medium sanddisplaying trough-cross stratification suggesting a north-ward-directed paleo-flow. This deposit is interpreted as anearshore deposit, and indicates a paleo-sea-level atleast 1.5m above the modern one. The upper beach-rockis observed above MSL, and is composed of tabularbeds of shell rich, poorly sorted conglomerate up to20 cm thick, interfingering fine to medium sand layersshowing a 4–61 dip, and with seaward inclined plane-parallel laminations. The dated shells mentioned abovealmost certainly originated from the conglomeratedeposit within the upper beach-rock, which is arrangedas a seaward inclined surface. We interpret this as alower beach-face deposit that indicates a sea level atleast 1.5m above present. Therefore, the strongest pieceof evidence sustaining secondary sea-level oscillationsactually indicates the opposite, i.e., higher sea-levels atthe time of the oscillations. In addition, because the

ructures and sediment textures (after Flexor and Martin, 1979).


shells are reworked, they can only indicate the oldestpossible age for the formation of the sedimentarydeposit (Lessa and Angulo, 1998).

4.3.3. Evidence of lower sea-level derived from vegetal

detritus embedded in beach deposits

Martin and Suguio (1975) have found evidence forthe 4100–3800 yr BP (4200–3700 cal yr BP) oscillationin a 2–3 cm thick accumulation of plant debris at Barrade Icapara (Figs. 1c and 15a). This debris, dated at33707100 yr BP (Gif-3430, 3857–3382 cal yr BP), wasinterpreted as part of an upper-beach sediment deposit,at a level close to a berm crest. The reasoning behind theinterpretation was that debris normally accumulatesclose to the high-water mark. Because the samples werecollected 0.3m above the present high water level, thepaleo-sea-level would have been 0.370.2m (Suguioet al., 1976), 0.570.3m (Suguio and Martin, 1978) or0.670.3m (Suguio et al., 1980).

The plant debris was covered by a 3m thick layer ofsand with plane-parallel, sub-horizontal stratificationand low-angle cross-stratification characteristic of beachface. In accordance with Martin and Suguio (1975), thesurface associated with the detritus would have been

(a)

(b)

(c)

(d)

Fig. 15. Schematic cross sections of Barra de Icapara south of state of

Sao Paulo: (a) present stage, (b) plant debris deposited during the

transgressive period, (c) plant debris fossilization stage and (d) present

erosion of Holocene terrace (after Martin and Suguio, 1975). For

location see Fig. 1c.

buried by foreshore sands in the following transgression(Figs. 15b and c), and then capped by the upper part ofthe sand sequence during the subsequent regression(Fig. 15d). This evolutionary model is rather implausiblegiven the erosion associated with the ravinement surfaceduring a transgressive episode. It is more likely, basedon the sedimentary structures we have observed in theoutcrop at the site, that these debris were depositedsomewhere in the lower half of the beach face, andare therefore indicative of a higher, and not a lower,paleo-sea-level. In addition, as with the shells from thebeach-rock, the dated samples are reworked and onlyindicate the oldest possible age for the formation of thesedimentary deposit.

4.3.4. Evidence of lower sea-level derived from ‘‘mangrove’’

swamp deposit

Another evidence for the 4100–3800 yr BP (4200–3700 cal yr BP) oscillation is a mangrove swamp depositfrom Rio do Fogo region in NE Brazil (Fig. 1a) (Martinet al. (2003). A layer of muddy sediments with woodfragments and shells of Lucina pectinata outcrops on thebeach face between low- and high-tide levels. Thedeposit, positioned slightly below the upper surface ofmodern mangrove swamp, was interpreted as a man-grove deposit and a relative sea-level slightly below thepresent level at the time of deposition. A wood fragmentfrom this outcrop was dated at 4200–4096 cal yr BP(Martin et al., 2003).

Bezerra et al. (2003) suggest that, between 4200 and2100 cal yr BP, sea level was continuously low (2 to�1m) (Fig. 16), along a 200 km long stretch of coastlinein Rio Grande do Norte (including Rio do Fogo). Asample of peat apparently obtained from the sameoutcrop visited by Martin et al. (2003) was dated at3550760 yr BP (BETA-121263, 3580–3320 cal yr BP),and would indicate a paleo-sea-level of 0.271.0m.Bezerra et al. (2003) argue that the paleo-sea-levelbehavior during the late Holocene in this coast has beenlargely influenced by tectonics and possibly wind-wavepatterns. Therefore, the peat (or mangrove deposit byMartin et al., 2003) deposit cannot be used to endorsethe existence of high-frequency oscillations.

4.3.5. Evidence of paleo-sea-levels higher than present at

the time of the suggested regressions (4200– 3700 and

2700– 2100 cal yr BP)

Several paleo-sea level indicators pointing to sea levelhigher than present in the intervals of the proposedhigh-frequency oscillations, have been published in theliterature (Table 1 and Fig. 17). Forty-six in situ, morereliable indicators are carbonate algae, corals, vermetidsand oysters, whilst 38 less reliable indicators are relatedto wood fragments and mollusk shells (normally A.

brasiliana) in paleo-estuarine sediments and plantdetritus and shells in barrier sediments (Table 1).

ARTICLE IN PRESS

Fig. 16. Sea-level envelops for two coastal sectors in the State of Rio Grande do Norte (after Bezerra et al., 2003).

Fig. 17. Distribution of more reliable paleo-sea-level indicators showing higher sea-levels during the time of the supposed negative high-frequency

oscillations: (i) algae, corals, oysters and vermetid that indicate a minimum paleo-sea-levels (triangle and dashed line); and (ii) vermetids that indicate

paleo-sea-levels with margin of error of 71.0m (squares and full line).


Fourteen paleo-sea-levels between 0.871.0 and4.471.0m are indicated for the interval 4200–3700 cal yrBP, whereas 21 paleo-levels between 0.871.0 and3.071.0m exist for the interval 2700–2100 cal yr BP.

In Cananeia, Ybert et al. (2003) found, by means ofdiatoms, that ‘‘the study area was occupied by a lagoonfrom at least 4900 up to ca. 3470 cal yr BP’’ when ‘‘sea-level was at least 1.2–2m higher than at present’’. Theregion subsequently underwent a ‘‘continuous fall of thesea-level’’.

Regarding the above discussion it is concluded thatnone of the paleo-sea-level indicators used to supportthe high-frequency sea-level oscillations, with ampli-tudes of more than 1m, are reliable or have beencorrectly interpreted. On the other hand, there areseveral reliable indicators that sustain a higher-thanpresent paleo-sea-level in the time interval of thesuggested oscillations.

5. Final remarks

The discussions above show that a significant part ofthe data set used to investigate crucial points of the

Brazilian sea-level history in the Holocene are problematicas sea-level indicators. The great majority of the dataanalyzed (amounting to about 70% of the data set) areeither non-conclusive or have been misinterpreted. Thishas led to several doubtful or erroneous paleo-sea-levelresults. At this point, we believe that a good approxima-tion of a general sea-level history can be best achievedwith the use of vermetids, which provide the most reliableestimate of age and elevation of former sea level.

Vermetid possesses several attributes that make themideal for paleo-sea-level reconstruction. Laborel (1979,1986) indicated that reefs of the vermetid Petaloconchus

(Macrophragma) varians could be used to assess paleo-sea-levels with a precision of 70.1 to 71.0m along theBrazilian coast, depending on the wave-energy level ofthe coast (higher waves imply in less precision). Anguloet al. (1999), after considering three other sources oferrors in the evaluation of the paleo-sea-level with fossilvermetid (remains may not correspond to the upperlimit of the formation, changes in the hydrodynamiccharacteristics of the coast through time, and referenceused to assess the vertical displacement of the vermetidreef), suggested that a precision of 71m should alwaysbe used.


Based on the existing data, it is not possible to suggestthat the elevation of the Holocene highstand has beensignificantly different along the Brazilian coast betweenPernambuco (81S) and Parana (261S), because themargins of error of the paleo-sea-level indicators arelarger than the suggested elevation differences. Aregional difference is only detectable in the south ofthe State of Santa Catarina (281S). Hence, when dealingwith the calibrated vermetid radiocarbon datings, datafrom the two regions were separated in order to defineregional trends (Fig. 18). Eight possible outliers wereidentified in Fig. 18 (samples Bah-1129, Bah-1152,Bah-1221, Bah-711, Bah-567, Bah-1586, Bah-1236 andGX-14061). These outliers plot outside the 2 standarddeviation envelope around the MSL trend for theirrespective regions, calculated by a 5th-order polynomial

-1

0

1

2

3

4

5

6

4567

1,000 ca

m

Fig. 18. Scatter of the vermetid samples dated along the Eastern Brazilian coa

indicators with vertical margins of error, showing an average paleo-sea-level t

empty squares and dashed lines, whereas the remaining data have empty circle

and squares.

Fig. 19. Sea-level envelop and paleo-sea-level reconstructions for the Brazilia

and circles), based on vermetid samples and with outliers removed.

(Fig. 18). With the removal of these outliers fromthe data set a new trend is then drawn for both regions(Fig. 19).

The highstand elevation does not appear to haveexceeded 4m to the north of Santa Catarina, wheremaximum Holocene sea level was about 2.1m. Thishighstand may have lasted for a couple of hundreds ofyears between 5000 and 5800 cal yr BP, without thedistinct peak (Fig. 4) that characterizes the sea-levelcurves in Martin et al. (2003). Similar elevations for thePSLM are indicated in South Africa (3.5m, Ramsay andCooper, 2002), northeastern Australia (1.7m, Larcombeet al., 1995), southeast Australia (2m, Baker et al., 2001)and south Australia (1–3m Belperio et al., 2002),suggesting that broadly similar hydro-isostaticadjustments (Chappell et al., 1982) may have occurred

0123

l.yr BP

st with their margin of error. A 5th-order polynomial was fit within the

rend. Data from the southern State of Santa Catarina are identified by

s and solid lines. Outliers within the data set are shown with full circles

n coast north 281 (solid line and squares) and south of 281 (dashed line


throughout the Southern Hemisphere, as predicted byglobal isostatic models (Clark et al., 1978; Lambeck andNakada, 1990; Fleming et al., 1998; Peltier, 1999; Milneet al., 2005). For the South American coast, modelresults by Milne et al. (2005) suggest that the highstandoccurred at 7000 cal yr BP, reaching �4m in Pernam-buco and Rio de Janeiro (Fig. 20a), and �2.5m in SantaCatarina (Fig. 20b). The model suggests a period ofrelative sea stabilization between 7000 and 5000 cal yrBP falling steadily after then. The time of the predictedhighstand departs from that suggested by the data(between 5000 and 5800 cal yr BP, Fig. 19). Thisdifference is ascribed to the presence of only one sampleolder than 6000 cal yr BP in the data set (Fig. 18), andcannot be taken as representative. If the sample isremoved, the predicted paleo-sea-level older than5800 cal yr BP becomes a fine extension of the observedtrend (Fig. 20).

It is not possible to affirm that the MSL-crossing timeoccurred at different times along the coast. Informationobtained from vermetids allows us to suggest that theMSL-crossing time occurred as late as 6500 cal yr BP.On the other hand, information derived from wooddebris and shell samples embedded in coastal-plainsediments, indicate that the MSL-crossing time tookplace as early as 7550 cal yr BP. This last date agreeswell with the crossover time predicted by Milne et al.(2005) (Fig. 20), which we suggest can be taken as areference for the relative sea-level variation in the last10,000 cal yr BP.

There is no reason to assume any Holocene change ofthe geoid in eastern Brazil, as advocated by Martin et al.(1985) and implied by Martin et al. (2003). Differencesbetween paleo-sea-level trends in Brazil only occur inareas undergoing localized tectonic subsidence, which isthe case of the State of Para, in north Brazil (Fig. 1a)(Behling et al., 2001; Cohen et al., 2005; Souza Filhoet al., in press), State of Rio Grande do Norte (Bezerraet al., 2003) and the surroundings of Baıa de Todos osSantos (Martin et al., 1984a; Carvalho, 2000). In the

-2

-1

0

1

2

3

4

5

6

012345678910

1,000 cal. yr BP

m

Pernambuco Rio de Janeiro

(a) (b

Fig. 20. Sea level envelopes for the region between Pernambuco and Parana (

sea-level behavior predicted by the geophysical simulations made by Milne e

State of Para, sea level may have never been higher thanthe current one throughout the Holocene (Souza Filhoet al. in press).

According with the data set, but differently to what issuggested by the geophysical simulation, sea levelapparently withstood an elevation between 2 and 3muntil about 2000 cal yr BP, as suggested by the envelopein Fig. 19 and by diatom data from Cananeia (Ybertet al., 2003). Coincidently, higher mid-Holocene sea-level was sustained up to 2500 yr BP in Australia assuggested by Chappell (1983), Larcombe et al. (1995)and Lessa and Masselink (2005). Uneven rates of sea-level fall in the late Holocene, with higher rates in thelast 2000 years, would have important implications tothe understanding of the evolution of the coastline, as itcould impact the rate of shore-normal sediment transferfrom the inner-shelf toward the shoreline.

The evidence for high-frequency (300–600 years) sea-level oscillations in the late Holocene are inconsistent, asinitially discussed by Angulo and Lessa (1997) and Lessaand Angulo (1998), and are contradicted by the wide-spread evidence of higher paleo-sea-levels in the timeperiod of the alleged oscillations (4200–3700 and2700–2100 cal yr BP) (Fig. 17). In Rio Grande do Norte,Bezerra et al. (2003) found no record of the oscillationsin the expected time frames, a disagreement that wasascribed to recent tectonism or changes in the wind/wave climate. In fact, the paleo-sea-level patternsobserved in two sectors of a 300 km long coastal zonediffer significantly (Fig. 16), and suggests importanttectonic controls (the area is well known for its seismicactivity—DeMets et al., 1990; Bezerra and Vita-Finzi,2000).

The secondary oscillations in Brazil are suggested tohave occurred along a latitudinal span of about 251,between Rio Grande do Norte in the north and Lagunain the south (Suguio et al., 1985; Martin et al., 2003).These oscillations, with an amplitude of 5m extendingover such a large area, would be characterized by riseand fall rates between 16 and 32mm/year, which is three

-1

0

1

2

3

4

5

6

012345678910

-2

1,000 cal. yr BP

m

Santa Catarina

)

a), and the southern State of Santa Catarina (b), plotted with the paleo-

t al. (2005).


times higher than the suggested rising rates (10mm/year) of the fastest sea-level rise period (15000–8000 calyr BP) in far-field sites (Fleming et al., 1998). Suchoscillations would be expected to leave records inneighboring regions, either to the north or south ofthe established limits. An inspection of other publishedcurves in South America shows that there is hardly anycoincidence with the paleo-sea-level pattern proposed inArgentina or in Suriname (Fig. 21).

Throughout the Southern Hemisphere there appearsto be a broadly consistent sea-level trend, with a smoothor gently oscillating decline of sea level after a Holocenemaximum between 2 and 3.5m. It cannot be ruled out,however, that the late Holocene sea-level fall in Brazilunderwent small-scale (few decimeters) oscillations, suchas those proposed by Baker et al. (2001) for SEAustralia. However, to test this hypothesis we need todevelop more precise indicators. A clear example of thelimited resolution of current approaches is that theyoungest samples of vermetids suggest paleo-sea-levelsbetween 0 and 1m in the last 500 years (Fig. 18).Although current MSL is within the margin of error ofall these samples, the mean position is clearly over-estimated. This may also be the case for the Surinamecurve in Fig. 21.

The proposition of secondary sea-level oscillationsalong the Brazilian coast in the late Holocene has causedproblems in the interpretation of the coastal geomor-

Fig. 21. Sea-level curves for the last 6000 years in the Southern Hemispher

sea-level after the highstand, modified from Isla (1989) and Angulo and Les

with calibrated years, the other curves with 14C years BP.

phology and the sedimentary record, especially due tothe premise that such oscillations would be able todevelop new strandplain or barrier/lagoon systems. Atthe time when the chronology of the coastal landscapewas first being uncovered, and in the light of the thenrecently proposed sea-level curves, some Brazilianresearchers mistakenly interpreted the Pleistocene bar-riers/strandplains as of Holocene age, associated withHolocene highstand. In the same way, the real Holocenehighstand barrier/strandplains were interpreted as re-sulting from a transgression at 3700 cal yr BP (Martinand Suguio, 1976; Maia et al., 1984; Muehe and Ignarra,1984; Neto, 1984; Perrin, 1984; Turcq et al., 1986;Muehe and Correa, 1988). More recently, Gomez et al.(2005) argued, on the basis of a single radiocarbon dateobtained from an underwater sedimentary deposit inBahıa Blanca estuary (Argentina, Fig. 1a), that largescale late Holocene sea-level oscillations also occurred inArgentina, thereby contradicting the majority of thestudies conducted in that country to date.

Although more than 1000 radiocarbon dates from theBrazilian coast exist and many hundreds of paleo-sea-level assessments have been performed, at present wecan say that the Brazilian Holocene sea-level history isstill far from complete. The current data set suggests aprogressive decline, with uneven rates, of the relativesea-level since the mid-Holocene highstand. Higherresolution studies of the sedimentary record (especially

e. All curves present similar elevations and show a gradual decline of

sa (1997). The curves from southeast Australia and Brazil were drawn


biogenic formations) are necessary to obtain a clearerpicture of the regional differences and possible smallscale oscillations as those found by Baker et al. (2001) inSE Australia. Perhaps even more important are in-vestigations on the continental shelf, due to a completelack of sound information on the sea-level history priorto 7000 years ago.

Acknowledgements

R.A. and G.L. are sponsored by a CNPq fellowship.M.C.S. is a CNPq scholarship holder. G.L. thanksFAPESB for traveling support during the preparation ofthis manuscript. This investigation has been supportedby the CNPq project 4710.42/2003-0. Drs. KenitiroSuguio and Louis Martin kindly provided informationon some samples they have used to determine thepublished paleo-sea-levels. We also would like to thankProfessor J. Rose and an anonymous referee for theircomments, which significantly contributed to improvethe quality of the manuscript.

Appendix A. Supplementary Materials

The online version of this article contains additionalsupplementary data. Please visit doi:10.1016/j.quascir-ev.2005.03.008.

Reference

Alber, M., 2002. A conceptual model of estuarine freshwater inflow

management. Estuaries 25, 1246–1261.

Almeida, F.F.M.de, 1955. Geologia e petrologia do arquipelago de

Fernando de Noronha. DNPM Divisao Geologia e Mineralogia.

Monografia 13 181.

Angulo, R.J., 1994. Indicadores morfologicos e sedimentologicos de

paleonıveis marinhos quaternarios na costa paranaense. Boletim

Paranaense de Geociencias 42, 185–202.

Angulo, R.J., Lessa, G., 1997. The Brazilian sea level curves: a critical

review with emphasis on the curves from Paranagua and Cananeia

regions. Marine Geology 140, 141–166.

Angulo, R.J., Suguio, K., 1995. Re-evaluation of the maxima of the

Holocene sea-level curve for the State of Parana, Brazil.

Palaeogeography, Palaeoclimatology, Palaeoecology 112, 385–393.

Angulo, R.J., Giannini, P.C.F., Suguio, K., Pessenda, L.C.R., 1999.

The relative sea-level changes in the last 5500 years southern Brazil

(Laguna-Imbituba region, Santa Catarina State) based on vermetid14C ages. Marine Geology 159, 327–339.

Angulo, R.J., Pessenda, L.C.R., de Souza, M.C., 2002. O significado

das datac- oes ao 14C na reconstruc- ao de paleonıveis marinhos e na

evoluc- ao das barreiras quaternarias do litoral paraense. Revista

Brasileira de Geociencias 32, 95–106.

Angulo, R.J., de Souza, M.C., Reimer, P., Sasaoka, S.K., 2005.

Reservoir effect of the southern and southeastern Brazilian coast.

Radiocarbon 47, 1–7.

Baker, R.G.V., Haworth, R.J., Flood, P.G., 2001. Warmer or cooler

late Holocene marine palaeoenvironments? Interpreting southeast

Australian and Brazilian sea-level changes using fixed biological

indicators and their d18O composition. Palaeogeography, Palaeo-

climatology, Palaeoecology 168, 249–272.

Behling, H., Cohen, M.C.L., Lara, R.J., 2001. Studies on Holocene

mangrove ecosystem dynamics of the Braganc-a Peninsula in north-

eastern Para, Brazil. Palaeogeography, Palaeoclimatology,

Palaeoecology 167, 225–242.

Belperio, A.P., Harvey, N., Bourman, R.P., 2002. Spatial and

temporal variability in the Holocene sea-level record of the South

Australian coastline. Sedimentary Geology 150, 153–169.

Bezerra, F.H.R., Vita-Finzi, C., 2000. How active is a passive margin?

Paleosismisity in northeastern Brazil. Geology 28, 591–594.

Bezerra, F.H.R., Barreto, A.M.F., Suguio, K., 2003. Holocene sea-

level history on the Rio Grande do Norte State coast, Brazil.

Marine Geology 196, 73–89.

Bigarella, J.J., 1946. Contribuic- ao ao estudo da planıcie litoranea do

Estado do Parana. Arquivos de Biologia e Tecnologia 1, 75–111.

Bigarella, J.J., 1950/51a. Contribuic- ao do estudo dos sambaquis no

Estado do Parana, 1 Regioes adjacentes as baıas de Paranagua e

Antonina. Arquivos de Biologia e Tecnologia 5–6, 231–292.

Bigarella, J.J., 1950/51b. Contribuic- ao ao estudo dos sambaquis no

Estado do Parana, I Regioes adjacentes a baıa de Guaratuba.

Arquivos de Biologia e Tecnologia 5–6, 293–321.

Bigarella, J.J., 1965. Subsıdios para o estudo das variac- oes do nıvel

oceanico no Quaternario Brasileiro. Anais da Academia Brasileira

de Ciencias 37, 263–278.

Bigarella, J.J., 1971. Variac- oes climaticas no Quaternario Superior do

Brasil e sua datac- ao radiometrica pelo metodo do Carbono 14.

Paleoclimas 1, 1–22.

Bigarella, J.J., Sanches, J., 1966. Contribuic- ao ao estudo dos

sedimentos praiais recentes. V—Praia suspensa do saco da

Tambarutaca, Municıpio de Paranagua—PR. Boletim Paranaense

de Geografia 18/20, 151–175.

Bigarella, J.J., Salamuni, R., Marques Filho, P.L., 1961. Metodo para

avaliac- ao do nıvel oceanico a epoca da formac- ao dos terrac-os de

construc- ao marinha. Boletim Paranaense de Geografia 4/5, 11–115.

Bittencourt, A.C.S.P., Martin, L., Vilas-Boas, G.S., Flexor, J.M.,

1978. Excursao sobre as formac- oes quaternarias do litoral da

regiao de Salvador, Bahia. In: International Symposium on Coastal

Evolution in the Quaternary, Sao Paulo, Brazil, Livro-Guia

(115pp).

Boehs,G., 2000. Ecologia populacional, reproduc- ao e contribuic- ao em

biomassa de Anomalocardia brasiliana (Gmelin, 1791) (Bivalvia:

Veneridae) na Baıa de Paranagua, Parana, Brasil. Unpublished

Master Thesis, Pos-Graduac- ao em Zoologia, Universidade Federal

do Parana (201pp).

Branner, J.C., 1902. Geology of the northeast coast of Brazil. Bulletin

of the Geological Society of America 13, 41–98.

Carvalho, J.B., 2000. Caracterizac- ao morfoestratigrafica do preen-

chimento sedimentar da Baıa de Iguape, Bahia: influencia das

variac- oes eustaticas do nıvel do mar e atividades tectonicas

recentes. Unpublished Master Thesis, Pos-Graduac- ao em Geolo-

gia, Universidade Federal da Bahia (120pp).

Chappell, J., 1983. Evidence for a smooth falling relative sea level to

north Queensland. Nature 302, 406–408.

Chappell, J., Rhodes, E.G., Thom, B.G., Wallensky, E., 1982. Hydro-

isostasy and the sea level isobase of 5500BP in north Queensland,

Australia. Marine Geology 49, 81–90.

Clark, J.A., Farrell, W.E., Peltier, W.R., 1978. Global changes in post-

glacial sea level: a numerical calculation. Quaternary Research 9,

265–287.

Cohen, M.C.L., Souza Filho, P.W.M., Lara, R.J., Behling, H.,

Angulo, R.J., 2005. A model of Holocene mangrove development

and relative sea-level changes on the Braganc-a Peninsula (Northern

Brazil). Wetland Ecology and Management in press.

Delibrias, C., Laborel, J., 1969. Recent variations of the sea level along

the Brazilian coast. Quaternaria 14, 45–49.

dx.doi.org/10.1016/j.quascirev.2005.03.008

dx.doi.org/10.1016/j.quascirev.2005.03.008


DeMets, C., Gordon, R.G., Argus, D.F., Stein, S., 1990. Current plate

motions. Geophysical Journal International 101, 425–478.

Dominguez, J.M.L., Bittencourt, A.C.S.P., Leao, Z.M.A.N.,

Azevedo, A.E.G., 1990. Geologia do Quaternario costeiro do

estado de Pernambuco. Revista Brasileira de Geociencias 20,

208–215.

Emperaire, J., Laming, A., 1956. Les sambaquis de la cote meridionale

du Bresil. Journal de la Societe des Americanistes, Nouvelle Serie

45, 1–166.

Fairbridge, R.W., 1961. Eustatic changes in sea level. In: Physics and

Chemistry of the Earth, vol. 4. Elsevier Science Ltd., London,

pp. 99–185.

Fish, S.K., DeBlasis, P.A.D., Gaspar, M.D., Fish, P.R., 2000. Eventos

incrementais na construc- ao de sambaquis, litoral sul do estado de

Santa Catarina. Revista do Museu de Arqueologia e Etnologia 10,

69–87.

Fleming, K., Johnstone, P., Zwartz, D., Yokoyama, Y., Lambeck, K.,

Chappell, J., 1998. Refining the eustatic sea-level curve since the

Last Glacial Maximum using far- and intermediate-field sites.

Earth and Planetary Science Letters 163, 327–342.

Flexor, J.M., Martin, L., 1979. Sur l’utilisation des gres coquilliers

de la region de Salvador (Bresil) dans la reconstruction des

lingnes de rivages Holocenes. In: Suguio, K., Fairchild, T.R.,

Martin, L., Flexor, J.M. (Eds.), Proceedings of the Inter-

national Symposium on Coastal Evolution in the Quaternary,

pp. 343–355.

Gaspar, M., 2000. Sambaqui: arqueologia do litoral brasileiro. Jorge

Zahar Ed., Rio de Janeiro (89pp).

Gaspar, M.D., 1998. Considerations of the sambaquis of the Brazilian

coast. Antiquity 72, 592–615.

Gaspar, M.D., Fish, P., Scheel-Ybert, R., Figuti, L., Klokler, D.,

Kneip, A., Ribeiro, L.B., Farias, D.S.E.D., Afonso, M., Karl, R.J.,

Eggers, S., Fish, S.K., DeBlasis, P., 2002. Padrao de assentamento

e formac- ao de sambaquis: arqueologia e preservac- ao em Santa

Catarina. Revista de Arqueologia do IPHAN 1, 57–62.

Genz, F., Lessa, G.C., Cirano, M., Lima, G.M.P.L., 2003. As

tendencias de longo prazo das vazoes fluviais no litoral do Estado

da Bahia. In: Anais do 91 Congresso da Associac- ao Brasileira de

Estudos do Quaternario, pp. 95–98.

Gomez, E.A., Martınez, D.E., Borel, C.M., Guerstein, G.R.,

Cusminsky, G.C., 2005. Negative sea level oscillation in Bahıa

Blanca estuary related to a global climatic change around

2650 yrBP. Journal of Coastal Research, SI39 in press.

Hartt, C.F., 1870. Geology and Physical Geography of Brazil. Fields,

Osgood & Co., Boston (620pp).

Hurt, W.R., 1974. The interrelationships between the natural

environment and four sambaquis, coast of Santa Catarina, Brazil.

Occasional Papers and Monographs, Indiana University Museum

1, 1–23.

Isla, F.I., 1989. Holocene sea-level fluctuation in the Southern

Hemisphere. Quaternary Science Reviews 8, 359–368.

Laborel, J., 1969. Les peuplements de madreporaires des cotes

tropicales du Bresil. Annales de l’Universite D’Abidjan, Serie E-

II Fascicule 3, Ecologie (261pp).

Laborel, J., 1977. Are reef-building Vermetids disappearing in the

South Atlantic? In: Proceedings of the 31 International Coral Reef

Symposium, pp. 233–237.

Laborel, J., 1979. Fixed marine organisms as biological indicator for

the study of recent sea level and climatic variations along the

Brazilian tropical coast. In: Suguio, K., Fairchild, T.R., Martin, L.,

Flexor, J.M. (Eds.), Proceedings of the International Symposium

on Coastal Evolution in the Quaternary, pp. 193–211.

Laborel, J., 1986. Vermetid gastropods as sea-level indicators. In: Van

de Plassche, O. (Ed.), Sea-Level Research: a Manual for

the Collection and Evaluation of Data. Geo Books, Norwich,

pp. 281–310.

Lambeck, K., Nakada, M., 1990. Late Pleistocene and Holocene sea

level change along the Australian coast. Palaeogeography, Palaeo-

climatology and Palaeoecology 89, 143–176.

Lamego, A.R., 1940. Restingas na costa do Brasil. Rio de Janeiro.

Boletim Divisao de Geologia e Mineralogia 96, 1–63.

Laming-Emperaire, A., 1968. Missions Archeologiques franc-aises au

Chili austral et au Bresil meridional: datations de quelques sites

par le radiocarbone. Journal de la Societe des Americanistes 57,

77–99.

Larcombe, P., Carter, R.M., Dye, J., Gagan, M.K., Johnson, D.P.,

1995. New evidence for episodic post-glacial sea-level rise, central

Great Barrier Reef, Australia. Marine Geology 127, 1–44.

Lessa, G.C., Angulo, R.J., 1998. Oscillations or not oscillations, that is

the question—reply. Marine Geology 150, 189–196.

Lessa, G.C., Masselink, G., 2005. Evidence of a mid-Holocene sea

level highstand form a sedimentary record of a macrotidal barrier

and paleoestuarine system in northwestern Australia. Journal of

Coastal Research.

Maack, R., 1947. Breves notıcias sobre a geologia dos Estados do

Parana e Santa Catarina. Arquivos de Biologia e Tecnologia 2,

66–154.

Maack, R., 1949. Espessura e sequencia dos sedimentos quaternarios

no litoral do Estado do Parana. Arquivos de Biologia e Tecnologia

44, 271–295.

Maia, M.C.A.C., Martin, L., Flexor, J.M., Azevedo, A.E.G., 1984. In:

Anais do 331 Congresso Brasileiro de Geologia, Sociedade

Brasileira de Geologia, vol. 1, pp. 105–118.

Martin, L., Suguio, K., 1975. The state of Sao Paulo coastal marine

Quaternary geology—the ancient strandlines. Anais da Academia

Brasileira de Ciencias 47, 249–263.

Martin, L., Suguio, K., 1976. O Quaternario marinho do

litoral do Estado de Sao Paulo. In: Anais do 291 Congresso

Brasileiro de Geologia, Sociedade Brasileira de Geologia, vol. 1,

pp. 281–294.

Martin, L., Suguio, K., 1978. Excursion route along the coastline

between the town of Cananeia (state of Sao Paulo) and Guaratiba

outlet (state of Rio de Janeiro). In: International Symposium on

Coastal Evolution, vol. 2. Special Publication, pp. 1–98.

Martin, L., Suguio, K., 1989. Excursion route along the Brazilian coast

between Santos (SP) and Campos (RJ) (North of State of Rio de

Janeiro). In: International Symposium on Global Changes in South

America During the Quaternary, vol. 2. Special Publication,

pp. 1–136.

Martin, L., Suguio, K., 1992. Variation of coastal dynamics during the

last 7000 years recorded in beach-ridge plains associated with river

mouths: example from the central Brazilian coast. Palaeogeogra-

phy, Palaeoclimatology, Palaeoecology 99, 119–140.

Martin, L., Flexor, J.M., Vilas-Boas, G.S., Bittencourt, A.C.S.P.,

Guimaraes, M.M.M., 1979a. Courbe de variation du niveau relatif

de la mer au cours des 7000 dernieres annees sur un secteur

homogene du littoral bresilien (nord de Salvador—Bahia).

In: Suguio, K., Fairchild, T.R., Martin, L., Flexor, J.M. (Eds.),

Proceedings of the International Symposium on Coastal Evolution

in the Quaternary, pp. 264–295.

Martin, L., Suguio, K., Flexor, J.M., 1979b. Le Quaternaire marin du

littoral bresilien entre Cananeia (SP) et Barra de Guaratiba (RJ).

In: Suguio, K., Fairchild, T.R., Martin, L., Flexor, J.M. (Eds.),

Proceedings of the International Symposium on Coastal Evolution

in the Quaternary, pp. 296–331.

Martin, L., Suguio, K., Flexor, J.M., 1979c. Precautions sur

l’utilisation des datations par le radiocarbone pour la reconstruc-

tion des anciens niveaux marins (Additif et correctif a: ‘‘Le

quaternaire marin du litoral bresilien entre Cananeia (SP) et Barra

de Guaratiba (RJ)’’. In: Suguio, K., Fairchild, T.R., Martin, L.,

Flexor, J.M. (Eds.), Proceedings of the International Symposium

on Coastal Evolution in the Quaternary, pp. 332–342.


Martin, L., Vilas-Boas, G.S., Flexor, J.M., 1979d. Introduc- ao ao

estudo do Quaternario do litoral do Estado da Bahia—Trecho

Salvador Ilheus. Revista Brasileira de Geociencias 9, 309–320.

Martin, L., Suguio, K., Flexor, J.M., Bittencourt, A.C.S.P., Vilas-

Boas, G.S., 1979/1980. Le quaternaire marin bresilien (littoral

pauliste, sud fluminense et bahianais). Cahiers O.R.S.T.O.M., Serie

Geologie 11, 95–124.

Martin, L., Bittencourt, A.C.S.P., Flexor, J.M., Vilas-Boas, G.S.,

1984a. Evidencia de um tectonismo quaternario nas costas do

Estado da Bahia. In: Anais do 331 Congresso Brasileiro de

Geologia, Sociedade Brasileira de Geologia, vol. 1, pp. 19–35.

Martin, L., Flexor, J.M., Blitzkow, D., Suguio, K., 1985. Geoid

change indications along the Brazilian coast during the last 7000

years. In: Proceedings of the Fifth International Coral Reef

Congress, vol. 3, pp. 85–90.

Martin, L., Suguio, K., Flexor, J.M., 1986a. Relative sea-level

reconstruction during the last 7000 years along the States of

Parana and Santa Catarina coastal plains: additional information

derived from shell-middens. Quaternary of South America and

Antarctic Peninsula 4, 219–236.

Martin, L., Suguio, K., Flexor, J.M., 1986b. Shell middens as a source

for additional information in Holocene shoreline and sea-level

reconstruction: examples from the coast of Brazil. In: Van de

Plassche, O. (Ed.), Sea-Level Research: a Manual for the

Collection and Evaluation of Data. Geo Books, Free University,

Amsterdam, pp. 503–521.

Martin, L., Suguio, K., Flexor, J.M., 1987. Flutuac- oes do nıvel

relativo do mar no Quaternario e seu papel na sedimentac- ao

costeira: Exemplos brasileiros. In: Simposio de Ecossistemas da

Costa Sul Sudeste Brasileira, vol. 1. Publicac- ao CIESP, pp. 40–61.

Martin, L., Suguio, K., Flexor, J.M., 1988a. Hauts niveaux marins

pleistocenes du litoral bresilien. Palaeogeography, Palaeoclimatol-

ogy, Palaeoecology 68, 231–239.

Martin, L., Suguio, K., Flexor, J.M., Azevedo, A.E.G., 1988b. Mapa

geologico do Quaternario costeiro dos Estados do Parana e Santa

Catarina. DNPM Serie Geologia 28, 1–40.

Martin, L., Suguio, K., Flexor, J.M., Dominguez, J.M.L., Bittencourt,

A.C.S.P., 1996. Quaternary sea-level history and variation in

dynamics along the central Brazil Coast: consequences on coastal

plain construction. Anais da Academia Brasileira de Ciencias 68,

303–354.

Martin, L., Suguio, K., Dominguez, J.M.L., Flexor, J.M., 1997.

Geologia do Quaternario costeiro do litoral norte do Rio de janeiro

e do Espırito Santo. CPRM Servic-o Geologico do Brasil (112pp).

Martin, L., Bittencourt, A.C.S.P., Dominguez, J.M.L., Flexor, J.M.,

Suguio, K., 1998. Oscillations or not oscillations, that is the

question: comment on Angulo R.J. and Lessa G.C. ‘‘The Brazilian

sea-level curves: a critical review emphasis on the curves from the

Paranagua and Cananeia regions’’ [Mar. Geol. 140, 141–166].


Martin, L., Dominguez, J.M.L., Bittencourt, A.C.S.P., 2003. Fluctu-

ating Holocene sea levels is eastern and southeastern Brazil:

evidence from a multiple fossil and geometric indicators. Journal of

Coastal Research 19, 101–124.

Milne, G.A., Long, A.J., Bassett, E., 2005. Modeling Holocene relative

sea-level observations from the Caribbean and South America.

Quaternary Science Reviews 24 (10–11), 1183–1202.

Muehe, D., Correa, C.H.T., 1988. Os ‘‘arenitos de restinga’’ do cordao

litoraneo da Mac-ambaba/Lagoa de Araruama—RJ. In: In: Anais

do 341 Congresso Brasileiro de Geologia, Sociedade Brasileira de

Geologia, vol. 2, pp. 553–561.

Muehe, D., Ignarra, S., 1984. Arenito de praia submerso em frente a

praia de Itaipuacu -RJ. In: Anais do 331 Congresso Brasileiro de

Geologia, Sociedade Brasileira de Geologia, vol. 2, pp. 341–346.

Nadal de Masi, M.A., 2001. Pescadores coletores da costa sul do

Brasil. Rio Grande do Sul, Pesquisas Antropologia 57, 1–136.

Neto, R.C., 1984. Algumas considerac- oes sobre a origem do sistema

lagunar de Araruama. In: Lacerda, L.D., et al. (Eds.), Restingas—

Origem, Estruturas, Processos. Universidade Federal Fluminense,

pp. 61–63.

Pekarova, P., Miklanek, P., Pekar, J., 2003. Spatial and temporal

runoff oscillation analysis of the main rivers of the world during

the 19th–20th centuries. Journal of Hydrology 274, 62–79.

Peltier, W.R., 1999. Global sea level rise and glacial isostatic

adjustment. Global and Planetary Change 20, 93–123.

Perrin, P., 1984. Evoluc- ao da costa fluminense entre as pontas de

Itacoatiara e Negra. Preenchimentos e restingas. In: Lacerda, L.D.,

et al. (Eds.), Restingas—Origem, Estruturas, Processos. Universi-

dade Federal Fluminens, pp. 65–73.

Ramsay, P.J., Cooper, J.A.G., 2002. Late Quaternary Sea-Level

Change in South Africa. Quaternary Research 57, 82–90.

Reineck, H.E., Singh, I.B., 1973. Depositional Sedimentary Environ-

ments. Springer, Berlin (439pp).

Roep, T.B., 1986. Sea-level markers in coastal barrier sands: examples

from the North Sea coast. In: Van de Plassche, O. (Ed.), Sea-Level

Research: a Manual for the Collection and Evaluation of Data.

Geo Books, Free University, Amsterdam, pp. 97–128.

Roy, P., Cowell, P.J., Ferland, M.A., Thom, B.G., 1995. Wave

dominated coasts. In: Carter, R.W.G., Woodroffe, C.D. (Eds.),

Coastal Evolution. Cambridge University Press, Cambridge,

pp. 121–186.

Short, A.D., 1984. Beach and nearshore facies: southeast Australia.


Souza, M.C., Angulo, R.J., Pessenda, L.C.R.de, 2001. Evoluc- ao

paleogeografica da planıcie costeira de Itapoa, litoral norte de

Santa Catarina. Revista Brasileira de Geociencias 31, 223–230.

Souza Filho, P.W.M., Cohen, M.C.L., Lara, R.J., Lessa, G.C., Koch,

B., Behling, H., 2005. Holocene coastal evolution and facies model

of the Braganc-a macrotidal flat on the Amazon Mangrove Coast,

Northern Brazil. Journal of Coastal Research.

Stuiver, M., Reimer, P.J., 1993. Extended 14C database and revised

CALIB radiocarbon calibration program. Radiocarbon 35,

215–230.

Suguio, K., Martin, L., 1978. Formac- oes quaternaria marinhas do

litoral paulista e sul fluminense. In: International Symposium on

Coastal Evolution in the Quaternary—the Brazilian National

Working Group for the IGCP; Project 61, vol. 1. Instituto de

Geociencias da Universidade de Sao Paulo, Special Publication,

pp. 1–55.

Suguio, K., Martin, L., Flexor, J.M., 1976. Les variations relatives du

niveau moyen de la mer au quaternaire recent dans la region de

Cananeia-Iguape, Sao Paulo. Boletim Instituto Geologico 7,

113–129.

Suguio, K., Martin, L., Flexor, J.M., 1980. Sea level fluctuations

during the past 6000 years along the coast of the state of Sao Paulo,

Brazil. In: Morner, N.A. (Ed.), Earth Rheology, Isostasy and

Eustasy. Wiley, Chichester, pp. 471–486.

Suguio, K., Martin, L., Bittencourt, A.C.S.P., Dominguez, J.M.L.,

Flexor, J.M., Azevedo, A.E.G.de, 1985. Flutuac- oes do nıvel

relativo do mar durante o quaternario superior ao longo do litoral

brasileiro e suas implicac- oes na sedimentac- ao costeira. Revista

Brasileira de Geociencias 15, 273–286.

Suguio, K., Martin, L., Flexor, J.M., Azevedo, A.E.G.de, 1986. The

Quaternary sedimentary deposits in the States of Parana and Santa

Catarina coastal plains. Quaternary of South America and

Antarctic Peninsula 2, 3–25.

Suguio, K., Martin, L., Flexor, J.M., 1992. Paleoshorelines and the

sambaquis of Brazil. In: Johnson, L.L., Stright, M. (Eds.),

Paleoshorelines and Prehistory: an Investigation of Method.

CRC Press, London, pp. 83–99.

Tomazelli, L.J., Villwock, J.A., Dillenburg, S.R., Bachi, F.A.,

Dehnhardt, B.A., 1998. Significance of present-day coastal erosion


and marine transgression, Rio Grande do Sul, southern Brazil.

Anais da Academia Brasileira de Ciencias 70, 221–229.

Turcq, B., Coe, R., Froidefond, J.M., 1986. Variability of beach ridges

on the coast of Marica (Rio de Janeiro, Brazil). Quaternary of

South America and Antarctic Peninsula 4, 45–57.

Van Andel, T.H., Laborel, J., 1964. Recent high sea-level stand near

Recife, Brazil. Science 145, 580–581.

Villwock, J.A., Tomazelli, L.J., Loss, E.L., Dehnhardt, E.A., Horn

Filho, N.O., Bachi, F.A., Dehnhardt, B.A., 1986. Geology of the

Rio Grande do Sul coastal province. Quaternary of South America

and Antarctic Peninsula 4, 79–97.

Wakamatsu, T., 1973. A ostra de Cananeia e seu cultivo. SUDELPA e

Instituto Oceanografico da Universidade de Sao Paulo (141pp).

Ybert, J.P., Bissa, W.M., Catharino, E.L.M., Kutner, M., 2003.

Environmental and sea-level variations on the southeastern

Brazilian coast during the Late Holocene with comments on

prehistoric human occupation. Palaeogeography, Palaeoclimatol-

ogy, Palaeoecology 189, 11–24.

Documents

A critical review of mid- to late-Holocene sea-level ... · A critical review of mid- to late-Holocene sea-level ﬂuctuations on the eastern Brazilian coastline Rodolfo J. Anguloa,,